Slavery in the sense of coercion of labor from people who are not free to choose their fate has a long history in China. For the most part, slaves came from various sources and were not always tied to race or ethnicity – although they sometimes were. Slaves could be peasants, commoners adopted as “sons” of noble families, eunuchs, Chinese or non-Chinese captured in wars, or later, small amounts of black slaves brought by an usually working under Europeans.
My specialty is in early Modern and modern China, so that’s where I’ll focus. Much of this comes from Pamela Crossley’s article “Slavery in Early Modern China” in the Cambridge World History of Slavery.
As Crossley says, “If the essential core of slavery is the physical coercion of labor from individuals who are invisible as legal persons, a good deal of China’s social history will come under the ‘slavery’ rubric”. At the same time, “there is no precise parallel to the Roman legal construction of slavery.
In China the absolute legal definition of slave status, or the associations with race and culture that might have inspired an equally absolute ideal of personal or national freedom, never emerged.” The same can be said of European/American style chattel slavery of the early modern and modern period.
There are a few reasons for this. First, in China, conceptions of property were traditionally more fluid and conditional than in Western legal traditions. “Chinese law and social institutions provided for instances of complete control by some people over others to whom they had no family relationship, people in China could not be reduced to res (a thing or object), because no res was defined in the law.” When the Ming empire discovered that Portuguese traders sold slaves as chattel Property, they forbade the sale of any Chinese to Portuguese traders – but did not forbid them from conducting their trade.
What kind of slavery existed? There was the ostensibly “voluntary” kind like concubinage, the aforementioned adoption of surrogate sons, and eunuchs. In practice while these groups could climb to great social heights, there were also laws proscribing their usurpation of the rights of nobles with a pedigree that did not include bondage.
It also included aspects of sexual slavery, in the case of concubines, and forced mutilation, in the case of eunuchs. These were typically Chinese although some emperors famously took foreign concubines, especially in the Tang Dynasty. There were also “bondsmen” to noble houses, a category above slaves who had restrictions on their actions and were not usually paid for their work.
There was also negotiated contracts a la “indentured servitude” that were legally negotiated in magistrate courts. Private sex slavery of women – prostitution – was the most common form of slavery throughout the empire.
Chinese law also differentiated people by “commoner” and “base” status. Commoners were born to their status and were generally more protected by laws than “base” people. Base status could be achieved by being born with congenital defects, being a prisoner of war or criminal, or by being identified as “idle”, in other words despondently poor.
As Crossley writes, “Poor people generally – and the base population specifically – performed the menial, nonagricultural tasks that were popularly despised. They guarded the fields, slopped night-soil, pounded earth (for building of walls and houses), gathered firewood, burned charcoal, and dug ditches and graves.” Only base status people could be enslaved by Chinese law. In times of war or large state projects, these restrictions were sometimes dropped, and mass corvee labor was the norm.
Crossley analyzes a customary wedding prayer from the early Qing period that illustrates the categories of slavery.
“Consistent with other documents for the period, the prayer carefully distinguishes between the status of house slaves (who are desired to be Chinese in this Chinese household) and farm and field slaves, who should be “foreigners.” Beautiful slaves (no gender specified) will take care of the entertainment, and as a final flourish, the link between perceived physical deformity and servility provides the punch line of the recitation:
Gold and silver to fill my coffers year after year,
Wheat and rice to fill my barns at every harvest.
Chinese slaves to look after these treasures,
Foreign slaves to tend my livestock,
Fleet-footed slaves to attend me when I ride,
Strong slaves to till the fields,
Beautiful slaves to strum the harp and fill my wine cup,
Slender-waisted slaves to sing and dance,
Midgets to hold the candle by my dining couch.”
What about the kunlun slaves from East Africa and South East Asia?
The phrase “kunlun” refers in classical Chinese discourse to black slaves, largely, as you say, from East Africa and Southeast Asia. Black slaves were known to ancient China through Arab and Persian traders, who sold and utilized these slaves throughout the maritime networks of Southern China and Southeast Asia.
Some elite Chinese officials even owned black slaves, although they would have not been considered “chattel” under Chinese legal conceptions of slavery. For example, sources from the time of the Liu Song dynasty (420-479) describe a “kunlun slave” who was ‘”often at [the emperor’s] side. He was ordered to beat the ministers and officials with a stick,” and even the highest-ranking ministers “feared his venom.”
The origin of the term “kunlun” (崑崙) is a bit unclear. As early as the Han dynasty, the Kunlun Mountains in northwest China were viewed as the home of the mythical Xi Wang Mu 西王母 (Queen Mother of the West). Over time, it took on a general meaning that sometimes connotated mystical and faraway lands and people.
The usage of Kunlun here seems unrelated to the original context. Kunlun was not exclusively used to describe black people either – there were areas called “kunlun areas” that contained people whose skin was not considered dark, and places where the inhabitants were described as dark-skinned but not called kunlun. There were also other words used to describe black slaves of the Arabs, like “sengchi/zengqi”, a transliteration of the Arabic word “zanj”, meaning dark. Quoting Winesky here:
“The Buddhist lexicographer Ruilin includes an entry on “The language of kunlun” in his dictionary Yiqie jing yinyi (The Sounds and Meanings of All the Scriptures), compiled between 783 and 820.28 Huilin uses the term kunlun as a category to describe dark-skinned people from the islands of the South Pacific:
‘Kunlun can also be written as gulun. They are the non-Chinese peoples from the east, those from the island states of the Southern Seas. Their bodies are black…. There are many types of them, including the zanj, the turmi, the kurdang, and the khmer. They are all base peoples. These countries lack ritual and propriety. They steal in order to live, and love to feed on humans for food, as if they were some sort of rakshas or a kind of evil ghost. The words they speak do not have any correct meaning at all…. They do extremely well when they enter the water, since they can stay there for a day without dying.’
Obviously, there are a lot of negative characteristics ascribed here – something that continues to this day. Although it would be a separate post, there is still racism in China w/r/t black people and Africans.
For example, just recently, it was reported that Africans living in China faced discrimination after being accused of secretly being the ones who spread the coronavirus due to their “dirty” habits. Not something we will go into much here, but we thought it would be remiss not to mention it.
However, this is not the whole story. By the time of the Tang Dynasty, mythical “kunlun slave” characters were popular parts of folklore. These characters usually had superhuman strength and ability, often serving as the hero of the story and displaying far more strength, virtue, and morality than those they were bound to.
For example, a famous character is Mo Juanhe, a black kunlun slave. The story begins with Mo’s mother, a poor women, having a dream about a “foreign monk” who presents her with a child with a shiny, black face, who he claims is her son and will one day achieve great power. The woman gives birth to a son with black skin “like iron.”
The boy later meets the Emperor, who was “delighted,” and said, “Where did we possibly get this kunlun boy?” The Emperor was excited that the boy’s surname, Mo (墨), which means “ink,” “corresponded with his appearance” and dubbed him “Mo Kunlun.” The Emperor also gave young Mo black clothes as a gift.
Winesky writer: “The Emperor’s automatic association of the word kunlun with the boy’s black skin assumes an audience familiar with the term. The non-Chinese monk and strangely “shiny” child evoke a sense of the supernatural. People with dark skin must have been exotic, based, on the Emperor’s jovial and surprised reaction to the boy. His gift of a black suit of clothes to match Mo’s black skin also suggests a sense of humor on the part of the Emperor.”
Again we will quote extensively from Winesky to convey the Mo story:
“The kunlun’s characteristic black skin also had magic powers, revealed in a Buddhist miracle tale in the Taiping guangji. A young slave girl, Xiao Jin, dreams of an old man mounted on a huge lion with reins held by two kunlun slaves. The old man, a Buddhist savior, tells Xiao Jin that he heard that demons were bothering her and traveled “ten thousand miles” to save her life. Xiao Jin asks the old man to relieve her terrible back and waist pain. The old man “ordered a kunlun to come forward and open his hand. He [the old man] rubbed his fingers on the palm of the kunlun’s hand… ” and became dyed “like black lacquer.” The old man then put his lacquered fingers on two moxibustion points on Xiao Jin’s back. When Xiao Jin awoke from her dream, her pain had ceased, and she began making Buddha statues and banners, evidence that this Buddhist story was written to encourage the general public to do good works. Although the old man is the Buddhist savior, it is the kunlun slave who has the supernatural power.”
Other stories depict heroes like Mo Le, who is the kunlun slave of a Chinese official who pines for the concubine of another. The concubine gives the official mysterious hand signals and sings strange songs in his presence, but the official cannot understand what she means. Seeing his master distraught, Mo Le asks him what is wrong, and is able to discern from the story that the concubine desires the official to save her from her captivity.
Mo valiantly rescues the girl, but must flee after the concubine’s former owner discovers what has happened. As Winesky says, “The portrayal of Mo Le as the tale’s hero places the kunlun slave in a positive light: his cunning, bravery, and sensitivity to the concubine far surpass that of his cowardly master. As he flies over walls with two grown people on his back and escapes from fifty soldiers, Mo Le’s physical prowess reaches mythical proportions. The author describes him as “like a winged bird, with speed like an eagle.”
In southern China especially, it was not unheard of to have black slaves. Zhu Yu writes about the practices of foreign traders and some rich households on the southern Chinese coast in his 1197 Notes on Pingzhou:
“Many of the wealthy households in Guangzhou raise devil slaves. They definitely have strength and can carry several hundred catties of weight. Their languages and preferences are not the same as ours, [but] their temperament is honest and they do not attempt to run away. They are also called wild people. Their coloring is black like ink, their lips are red and their teeth white, their hair is curly and yellowish. There are both females and males.
They were born in the mountains beyond the sea. They eat uncooked food. When they are captured, they are fed cooked food, which gives them diarrhea. This is called “changing their bowels.” This causes some of them to become sick and die. Those who do not die can be domesticated. Those who have been domesticated for a long time can understand human language, although they cannot speak it themselves. A type of wild man who comes from a place near the sea and who can enter the water without closing his eyes is called a kunlun slave.”
This discourse draws on the Chinese discourse of barbarian/civilized, or “cooked/uncooked” races, that imagines non-Chinese as barely human until/unless they accept Chinese language, ways, and customs. These accounts are therefore likely exaggerated in order to create this “ideal type” that is reflective of Chinese ideas about the self, rather than the reality of the “other”. In fact, many non-Chinese minoirites would be described in similar terms.
So in conclusion, black slaves were rare, but not unheard of, and mostly known through contacts with Arab traders and in southern China. They were not kept as chattel property, and were respectable and magical characters in some forms of folklore, but also were looked down upon as inferior in other contexts.
2. “The Magical Kunlun and “Devil Slaves”: Chinese Perceptions of Dark-skinned People and Africa before 1500”
Practices of slavery and servitude of Jurchen/Manchu during the time of the Qing empire (1636-1912)
Jurchen/Manchu society was built on slavery. This was something ingrained not only in their social structure but also their political rhetoric, with the relationship between ruler and subject conceptualized explicitly as being equivalent to that of a master and slave. That is not to say that all subjects of the Jurchen khan/Manchu emperor were slaves as we would understand them.
‘Slavery’ as conceptualized in Northeast Asia was not exactly chattel slavery, where an enslaved person became seen as transferrable property, but rather refers to a more general state of political and economic unfreedom whereby one individual became bound to another individual or group in effective perpetuity.
For that reason, we’re going to treat booi (‘bondservant’) status as being a form of slavery within that Northeast Asian conception. As we understand it, the difference between indentured servitude and slavery in a European context is a significant one, but in a Manchu context bonded servitude was part of the broader category of slavery.
In pre-conquest Jurchen society, we can distinguish three rough classes – at the top were ᡳᡵᡤᡝᠨ irgen, the village heads; these commanded the allegiance of ᠵᡠᡧᡝᠨ jušen, men who owned farmland; and finally there were ᠠᡥᠠ aha (slaves) and ᠪᠣᠣᡳ booi (bondservants), who were considered bonded in servitude to jušen masters.
As Pamela Crossley points out, though, the relationship between irgen and jušen could be considered in some ways comparable to that between jušen and aha, and a relationship of servitude became particularly apparent when, by the 1610s, one particular ᠪᡝᡳᠯᡝ beile (‘prince’), Nurgaci (aka Nurhaci), had achieved sufficient control to start considering himself ᡥᠠᠨ han (Khan) of the Jurchens. Nurgaci’s pronouncements reveal a paternalistic attitude from ruler to subject that corresponds with what we understand of prior Jurchen discourses as regards slavery: the enslaved person is conceptualised as a child in relation to the paternal figure of the master.
The origins of the aha and booi were various and not simply, as traditionally believed, exclusively Han Chinese war captives and their descendants. The booi companies of the Banner system appear to have been distinguished by ethnicity, and while Han Chinese were a significant part of them, there would have been ethnic Manchus, Mongols, and Koreans in the booi companies as well.
A major cause for the increase in number of enslaved people during the early part of Qing rule seems to have been economic desperation: many people sold themselves or their family members into slavery during the chaos of the Qing conquest in the 1640s-50s, and this pattern repeated in some cases during times of natural disasters: in the 1720s, there were reports of peasants around the city of Jingzhou selling themselves to the Banner garrison following a period of major flooding. But captives of all sorts were the most consistently enslaved by the Qing state: not just prisoners of war but also, at times, criminals.
The precise structure of Manchu society underwent massive changes after 1644, for the simple reason that most of the Manchu population relocated from the Northeast Asian plain to urban centres in China, a move which brought with it some significant alterations. To explain them, it is worth bringing up what the terms aha and booi were understood to mean.
Quoting directly from Evelyn Rawski’s The Last Emperors, ‘Whereas aha worked in fields, booi were in domestic service.’ The term booi itself highlights this: boo means ‘house[hold]’, and the -i suffix converts it to possessive form, thus a booi is ‘[someone] of the house[hold]’. The number of aha in the Qing state thus became far less significant (especially as the Bannermen came to sell off most of their agricultural allotments in the provinces), and most aha would in fact be ‘elevated’ up to booi. Ownership of people as household servants became an expected part of Manchu identity as part of the means by which they distinguished themselves from their Han neighbours.
Over the course of the Kangxi reign (1661-1720) the number of booi in Beijing remained relatively steady at around 230,000, although the number of non-booi in the capital Banners increased considerably from around 150,000 to around 385,000. One noticeable increase was that of Hanjun, whose numbers increased from around 75,000 to just under 200,000, a change second only to the increase of Manchus from around 50,000 to 150,000. Events later in the eighteenth century, however, show that the booi population was not atypically static compared to the ‘free’ Banner population.
Rather, many booi were working around the system. One of the great shocks that the Qianlong Emperor (r. 1735-96/9) faced was that he discovered that a substantial portion of his guards could not speak Manchu. While on the one hand perhaps reflective of declining language teaching standards (which he would go on to attempt to reverse), it seems the more significant issue (for the emperor at least) was that booi and other enslaved people in the Banenrs were exploiting loopholes and oversights. Most notably, many were having their children adopted by Manchu families, ensuring that even if they remained bonded, their children would end up in ‘free’ companies.
Some ‘entailed households’, which were run by booi patriarchs who had gained some degree of elevation due to service to the state (such as in battle), also came to claim status as ‘detached households’, a category used to refer to households of full Bannermen where the patriarch held no official post.
This state of affairs, unfortunately, did not last: by the 1750s the Banner administration began to much more actively monitor adoption processes and enforce the rule that Manchus were only to adopt Manchu orphans and that all orphaned Manchus had to be adopted by Manchus, irrespective of clan or sub-Banner affiliations, while the category of ‘entailed household’ was replaced with a more tightly monitored set of ‘separate-register households’.
From there on out, the bonded and enslaved members of the Banner system would remain as such until the end of the dynasty. There did remain, however, the option of outright escape, one with actually quite a high success rate. Coldo, the garrison commander at Xi’an in the early years of the Qianlong reign, reported that around 170-200 enslaved people escaped every year, of whom only 20 or so at most would ever be recaptured.
A final aspect I’d like to discuss, and one that must be treaded carefully, is the notion of Bannermen, even if not in the aha or booi, as ‘slaves’ of the emperor. We brought up the equivalencies of paternalistic rhetoric earlier, but the language used was very very clear: in memorials to the emperor, Bannermen referred to themselves in Manchu as ᠠᡥᠠ aha and in Chinese as 奴才 nucai, literally, ‘[this] slave’.
This is in contrast to Han Chinese officials, who used 臣 chen (‘[this] official’). Now, Bannermen were not chattel slaves, they were not considered the personal property of the emperor to buy and sell at his leisure. However, going back to the start again, within the Inner Asian conception of slavery as meaning a state of political and economic unfreedom, Bannermen were most certainly bound to the emperor’s will in a manner that most imperial subjects were not, and this rhetoric is not simply a bit of flourish, but a real reflection of how emperor and, presumably, Bannerman, conceptualised their relationship.
Mark Elliott highlights as a comparison the Janissary corps of the Ottoman Empire: relied upon to be effective in war and fiercely loyal, but whose status was clearly subordinated to their monarch, who reserved theoretically total control over their lives.
2. Pamela Crossley, A Translucent Mirror (1999)
3. Mark C. Elliott, The Manchu Way (2001)
4. Evelyn Rawski, The Last Emperors (1998)
5. Christine Moll-Murata, ‘Tributary Labour Relations in China During the Ming-Qing Transition’, IRSH 61 (2016)